Hello, thank you for the video. I want to ask one question about my validation part. Why did MOLUSCE say that my reference map (MODIS LULC) and simulated map (predicted using MOLUSCE) have different geometries? Because of this, I cannot validate my predicted map. Thank you for your help=)
@@gisremotesensing2639 Thank you so much. I made it. May I know what the current validation kappa is for? I used LULC 2002 and 2019 for 2036 LULC prediction. But I only got 0.53 validation kappa.
@@wanderinggeol current kappa validation is the kappa value for the ANN modeling. along with the value check the Neural Network learning curve whether it is overfit, underfit or good fit. increase and decrease the number of hidden layers to reach the optimal kappa value
Hello, i am getting this error "ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()" when I try update tables in Area Changes. Do you have any tip for this?
Sir, i have been using QGIS 3.18 Zurich for my thesis work, MOLUSCE plugin is not available in this version of QGIS. Kindly help me to use this plugin in QGIS 3.18
As MOLUSCE plugin is developed by Air Asia Survey and NextGIS and they didn't update the plugin for latest versions of QGIS it only works in 2.0.0 to 2.9.9 versions of QGIS
@@gisremotesensing2639 Thanks I almost figured that out from your paper !! Many of the other RUclips Videos are showing the year written in the file name determines the predicted year..!! lol. I have another question, for example we developed ANN to predict 2020 map using existing 2000 and 2010 maps and then validated the predicted 2020 map with the existing 2020 map. Arent we supposed to use the same model for predicting 2030, 2040 etc. Bcoz if I use 2010 and 2020 map to predict 2030 map, isn't that equivalent to dropping the model for which we have shown the validation in the previous step.
@@shobhitchaturvedi4866 Yeah model calibration is mandatory i.e. project the LULC for 2020 first and validate with the actual one. Following the same parameters use 2010 and 2020 to projectct 2030 and so on
@@gisremotesensing2639 Dear Sir, Thanks for your help.. I am facing difficulty using Molusce. Actually, I followed your instructions and used 2001 and 2011 LULC maps as inputs for the MLP and CA models to predict the 2021 LULC Map. However instead of predicting the 2021 Map, the predicted map looks 99.9% like 2011 map. I have looked at many moluce youtube videos and tried several times but this issue is not going anywhere. Can you suggest where I may be going wrong ? and how can i develop a correct future map.
I want to simulate 2018 using the land cover map of 2006 and 2012. However, the simulation map obtained gives almost the same result as 2012. Moreover, 2006 and 2012 are quite different from each other. How can I solve this problem? Also, I don't get any error while processing. I will be grateful if you help.
Try to use more spatial variables and change the parameters in transition potential modeling and cellular automata tab. for further guidance you may go through this paper. www.mdpi.com/2073-445X/10/6/584
Hi. Please I need help.. Hi guys, I need help. As part of my Final year project, I'm to do LULC prediction of 2045. I watch this video on how to use MOLUSCE plugin in QGIS.. but I'm getting the error message that " initial and spatial variable geometry are different".. please I need help
Iterations are the values randomly sampled from the input probability distributions and the outcome from that sample. It defines the sampled values from the input not for different time periods
Hello, I am having a check geometry error. My coordinate system is the same but the number of columns and rows is different. How do I do it all the same? Can you help me?
In order to fix the geometry issues use clip tool in ArcGIS or QGIS. remember that the cell size, columns and rows, extent and No Data Value must be same.
Hello, where did you get the data for GDP,slope. population? I've downloaded the LULC from USGS, but idk where to get the data in order to use the molusce
thank you sir but ask one quotation during MOLUSCE folder can not Crete change map say that An error has occurred while executing Python code: how to solve this error
@@gisremotesensing2639 An error has occurred while executing Python code: Traceback (most recent call last): File "C:/Users/power user/.qgis2/python/plugins\molusce\moluscetablewidget.py", line 49, in keyPressEvent data += self.item(r, c).text() + "\t" AttributeError: 'NoneType' object has no attribute 'text' this will show if ill do it...
Hi, thank you for this video. As part of my master's thesis I need to predict future land cover of my study area. Need some help with the input images. Will you be able to help?
@@gisremotesensing2639 I am facing two problems in using the MULUSCE plugin. 1) i am using sample data set that i downloaded. Geometry of all tiff are matching. Getting error while "Update Tables" under "Area Change" tab. 2) How to export my shape files (LULC and variables) to the TIFF with same geometry (Row, Column, No data value & Cell size). Would it be possible to converse directly to you in some way? Cant explain everything in RUclips comments. Thank you.
@@NishantDesaiAr 1- you are facing error 1 because you didn't classify your LULC data. 2- use "rasterize" tool in QGIS or "polygon to raster" tool in ArcGIS for conversion of shapefile data. 3- after rasterize your data you can use "Extraction" tool in QGIS or "Clip" or "Extract by Mask" tools in ArcGIS for fixing the geometry, extent, no data value etc.
Thank you for sharing it's really helpful. In neural network learning curve, how can I interprete the curve ? What's the standard range of min validation overall error and validation Kappa ?
There are three dynamics to observe the learning curves: underfit, overfit, and good fit. underfit curves are flat lines (a big gap b/w training and validation line) that means the model is unable to learn the training dataset. Overfit curves: the training line continues to decrease and the validation line decreases first but at a point it increases again So at starting there is not a significant gap between training and validation line but after some points the gap between the two Lines increases and such phenomenon illuminates that the model learned too much including fluctuations and noise etc. sometimes it also occurs if the model is trained for a long time. Good Fit curves: training and validation Lines move together and with a minimal gap between the two lines e.g. the Lines in this video. As ANN is a machine learning algorithm so the accuracy and kappa value above 0.9 considers good
I used qgis 2.18.28, input rasters and geometries of the rasters are matched. When I update the tables in area changes arising a python error: an error has occured while executing python code Please can you tell a method to solve that problem
@@gisremotesensing2639 I have clipped my AOI from DEM map and rasterized roads and settlements data using the same shp for my study area but geometry is still different. Thanks for your assistance.
![I.N "HALLUCINATION" | [Stray Kids : SKZ-PLAYER]](http://i.ytimg.com/vi/n5B5q1Hwt_U/mqdefault.jpg)
Sir, how does the algorithm know which year data you wish to predict?
Good Question.. It seems we are giving a file name with 2040 and that's it..seems unlikely..!!
Hello, thank you for the video. I want to ask one question about my validation part. Why did MOLUSCE say that my reference map (MODIS LULC) and simulated map (predicted using MOLUSCE) have different geometries? Because of this, I cannot validate my predicted map. Thank you for your help=)
How do I make the layer to have the same geometry. I've tried resampling, the rows and columns are the same but no data value is different
ruclips.net/video/QcpCNLrP1gI/видео.html
This is amazing ,which version of qgis is suitable for molusce plugin
Hello! MOLUSCE 4 has been released, now with support for the latest versions of QGIS
@@nextgis_channel thank you
Hi, how do know we are doing for the 2030 prediction because I'm not seeing you specifically point it out in the video?
we are using a ten year interval i.e. 2010 for intial year and 2020 for final year so the prediction should be for 2030
Informative and worth watching .
Thank you for sharing knowledge
Hi! What change detection analysis did you use in the MOLUSCE plugin?
Thanks a lot,
how to solve this problem LULC Change in Qgis . Geometries of the initial raster and final raster are different!
The geometries of my DEM, aspect, slope, distance from road, etc. maps don't match with my LULC maps. What can I do to match them?
ruclips.net/video/QcpCNLrP1gI/видео.html&ab_channel=GIS%26RemoteSensing
@@gisremotesensing2639 Thank you so much. I made it. May I know what the current validation kappa is for? I used LULC 2002 and 2019 for 2036 LULC prediction. But I only got 0.53 validation kappa.
@@wanderinggeol current kappa validation is the kappa value for the ANN modeling. along with the value check the Neural Network learning curve whether it is overfit, underfit or good fit. increase and decrease the number of hidden layers to reach the optimal kappa value
Hi! How were you able to convert the population data to raster? I can't seem to figure out the process. Thank you!
Hello, i am getting this error "ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()" when I try update tables in Area Changes. Do you have any tip for this?
Check the pixel type of your data. It must be integer
Sir, i have been using QGIS 3.18 Zurich for my thesis work, MOLUSCE plugin is not available in this version of QGIS. Kindly help me to use this plugin in QGIS 3.18
As MOLUSCE plugin is developed by Air Asia Survey and NextGIS and they didn't update the plugin for latest versions of QGIS it only works in 2.0.0 to 2.9.9 versions of QGIS
Thanks a lot Sir
Can you please make a video where you will consider the years for the spatial variables?
Thanks for this infomrative video.. I have a doubt.. how does Molusce know which year's LULC map we are looking to predict ?
The temporal gap between initial and final LULC determines the projected year
@@gisremotesensing2639 Thanks I almost figured that out from your paper !! Many of the other RUclips Videos are showing the year written in the file name determines the predicted year..!! lol.
I have another question, for example we developed ANN to predict 2020 map using existing 2000 and 2010 maps and then validated the predicted 2020 map with the existing 2020 map. Arent we supposed to use the same model for predicting 2030, 2040 etc.
Bcoz if I use 2010 and 2020 map to predict 2030 map, isn't that equivalent to dropping the model for which we have shown the validation in the previous step.
@@shobhitchaturvedi4866 Yeah model calibration is mandatory i.e. project the LULC for 2020 first and validate with the actual one. Following the same parameters use 2010 and 2020 to projectct 2030 and so on
@@gisremotesensing2639 Dear Sir, Thanks for your help.. I am facing difficulty using Molusce. Actually, I followed your instructions and used 2001 and 2011 LULC maps as inputs for the MLP and CA models to predict the 2021 LULC Map. However instead of predicting the 2021 Map, the predicted map looks 99.9% like 2011 map. I have looked at many moluce youtube videos and tried several times but this issue is not going anywhere.
Can you suggest where I may be going wrong ? and how can i develop a correct future map.
I want to simulate 2018 using the land cover map of 2006 and 2012. However, the simulation map obtained gives almost the same result as 2012. Moreover, 2006 and 2012 are quite different from each other. How can I solve this problem? Also, I don't get any error while processing. I will be grateful if you help.
Try to use more spatial variables and change the parameters in transition potential modeling and cellular automata tab. for further guidance you may go through this paper. www.mdpi.com/2073-445X/10/6/584
@@gisremotesensing2639 thank you for your help.
using MOLUSCE plugin in Qgis should I validate with any model to get know the data in molusce is okay?
Could you elaborate your question please
Hi. Please I need help.. Hi guys, I need help. As part of my Final year project, I'm to do LULC prediction of 2045. I watch this video on how to use MOLUSCE plugin in QGIS.. but I'm getting the error message that " initial and spatial variable geometry are different".. please I need help
ruclips.net/video/QcpCNLrP1gI/видео.html
In cellular automata simulation, if you change the simulation iteration to 2, for what year will your simulation be? Thank you
Iterations are the values randomly sampled from the input probability distributions and the outcome from that sample. It defines the sampled values from the input not for different time periods
you may go through this article for further understanding www.mdpi.com/2073-445X/10/6/584/htm
In my qgis MOLUSCE plugin is not available. What to do ?
ruclips.net/video/N8-G6IY3ueA/видео.html
Hi thank you for sharing great video, i want ask some question how to increase accuracy validation kappa? Bcs im stuck in 66%
Hello, I am having a check geometry error. My coordinate system is the same but the number of columns and rows is different. How do I do it all the same? Can you help me?
In order to fix the geometry issues use clip tool in ArcGIS or QGIS. remember that the cell size, columns and rows, extent and No Data Value must be same.
Hello, where did you get the data for GDP,slope. population? I've downloaded the LULC from USGS, but idk where to get the data in order to use the molusce
you can get the population and GDP data from local authorities and calculate slope from DEM
thank you sir but ask one quotation during MOLUSCE folder can not Crete change map say that An error has occurred while executing Python code: how to solve this error
use an older version of QGIS
sir how to copy the transition matrix to excel?
use your keyboard to copy and paste
@@gisremotesensing2639 what command sir..if ill input ctl copy it wont work..
@@joeparescario7823 select the table first and use ctrl+c and ctrl+v
@@gisremotesensing2639 An error has occurred while executing Python code:
Traceback (most recent call last):
File "C:/Users/power user/.qgis2/python/plugins\molusce\moluscetablewidget.py", line 49, in keyPressEvent
data += self.item(r, c).text() + "\t"
AttributeError: 'NoneType' object has no attribute 'text'
this will show if ill do it...
Hi, thank you for this video. As part of my master's thesis I need to predict future land cover of my study area. Need some help with the input images. Will you be able to help?
Hello. feel free to ask if you need help
@@gisremotesensing2639 I am facing two problems in using the MULUSCE plugin. 1) i am using sample data set that i downloaded. Geometry of all tiff are matching. Getting error while "Update Tables" under "Area Change" tab. 2) How to export my shape files (LULC and variables) to the TIFF with same geometry (Row, Column, No data value & Cell size).
Would it be possible to converse directly to you in some way? Cant explain everything in RUclips comments.
Thank you.
@@NishantDesaiAr
join our facebook page:
facebook.com/groups/391073008774787
@@NishantDesaiAr 1- you are facing error 1 because you didn't classify your LULC data. 2- use "rasterize" tool in QGIS or "polygon to raster" tool in ArcGIS for conversion of shapefile data.
3- after rasterize your data you can use "Extraction" tool in QGIS or "Clip" or "Extract by Mask" tools in ArcGIS for fixing the geometry, extent, no data value etc.
Thanks, dear, The video is very fine. But I am trying to install (molusc) in QGIS but can't install it. Do you help me, anybody, to solve the problem?
ruclips.net/video/N8-G6IY3ueA/видео.html&ab_channel=GIS%26RemoteSensing
Thank you for sharing it's really helpful. In neural network learning curve, how can I interprete the curve ?
What's the standard range of min validation overall error and validation Kappa ?
There are three dynamics to observe the learning curves: underfit, overfit, and good fit. underfit curves are flat lines (a big gap b/w training and validation line) that means the model is unable to learn the training dataset. Overfit curves: the training line continues to decrease and the validation line decreases first but at a point it increases again So at starting there is not a significant gap between training and validation line but after some points the gap between the two Lines increases and such phenomenon illuminates that the model learned too much including fluctuations and noise etc. sometimes it also occurs if the model is trained for a long time. Good Fit curves: training and validation Lines move together and with a minimal gap between the two lines e.g. the Lines in this video.
As ANN is a machine learning algorithm so the accuracy and kappa value above 0.9 considers good
@@gisremotesensing2639 Thank you so much for helping and your guidance.
I used qgis 2.18.28, input rasters and geometries of the rasters are matched. When I update the tables in area changes arising a python error: an error has occured while executing python code
Please can you tell a method to solve that problem
check the pixel type of your data. it must be 'integer'
@@gisremotesensing2639 thank you
@P20UP020 PUJA Use "Copy Raster" tool in ArcGIS
How do I get my raster files to match geometry?
use clip tool in ArcGIS or QGIS to fix the geometry issues
@@gisremotesensing2639 I have clipped my AOI from DEM map and rasterized roads and settlements data using the same shp for my study area but geometry is still different. Thanks for your assistance.
@@carolinekimani1689 check the cell size of your data
@@gisremotesensing2639 Thank you. I have successfully matched cell size and no data value but where do I change rows and columns?
@@carolinekimani1689 Use clip tool in ArcGIS or Extraction tool in QGIS
Hello, I need help
Hello. feel free to ask
@@gisremotesensing2639 Can I send you private message? Thank you
@@geadionio2190 drop you email, will get back to you
sir, can i get your contact details my masters project is stuck because of this .
Please drop your email here, I will get back to you
www.mdpi.com/2073-445X/10/6/584/htm